Skip to main content

How to configure Link Aggregation with LACP in Debian 12 or Ubuntu 24.04

Stephanie R.
Updated

Learn how to bond your interfaces to use them for link aggregation in Debian 12 (or newer) / Ubuntu 24.04 (Netplan configuration).

Link Aggregation Control Protocol (LACP) technology is designed to increase your server’s availability and boost the efficiency of your network connections. You can aggregate your network cards and make your network links redundant. This means that if one link goes down, traffic is automatically redirected to another available link. The available bandwidth is also doubled thanks to aggregation.

NOTE: While Debian 12 and newer images provided by OVHcloud utilize Netplan by default, there are two key exceptions where ifupdown (/etc/network/interfaces) is used instead:

  • Rescue mode: Although based on Debian 12, the rescue environment relies on the ifupdown utility.
  • Custom images: Debian installations performed using your own image may still use ifupdown for networking.

If you wish to configure link aggregation in rescue mode, or on a custom OS relying on ifupdown, please refer to this guide instead.

 

Requirements

OVHcloud Control Panel Access

  • Direct link: Dedicated Servers
  • Navigation path: Bare Metal Cloud > Dedicated servers > Select your server

 

Instructions

The values (MAC addresses, IP addresses, etc.) shown in the configurations and examples below are provided as examples. Of course, you must replace these values with your own.

Retrieving MAC addresses

Switch to the Network Interfaces tab and take note of the MAC addresses for each interface (public/private), which are displayed at the bottom of the menu.

dedicated_networkInterfacesTab-2.PNG

NOTE: The MAC address of the main public interface is the one receiving DHCP offers, both in the server's operating system and in rescue mode. This interface handles public connectivity in the default configuration.

Now that you know which MAC addresses are associated with each type (public/private) of interface, you need to retrieve the interface names.

 

Retrieving interfaces names

If you lose network connection to your server, follow the "Open KVM" steps from this guide.

To retrieve the names of the interfaces, execute the following command:

ip a

NOTE: This command will yield numerous interfaces. If you are having trouble determining which ones are your physical interfaces, the first interface will still have the server's public IP address attached to it by default.

Here's an output example:

1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN group default qlen 1000
    link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
    inet 127.0.0.1/8 scope host lo
       valid_lft forever preferred_lft forever
    inet6 ::1/128 scope host noprefixroute
       valid_lft forever preferred_lft forever
2: ens22f0np0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc mq state UP group default qlen 1000
    link/ether a1:b2:c3:d4:e5:c6 brd ff:ff:ff:ff:ff:ff
    inet 203.0.113.1/32 metric 100 scope global dynamic ens22f0np0
       valid_lft 71613sec preferred_lft 71613sec
    inet6 2001:db8:1:1b00:203:0:112:0/56 scope global
       valid_lft forever preferred_lft forever
    inet6 fe80::a6b2:c3ff:fed4:e5c6/64 scope link
       valid_lft forever preferred_lft forever
3: ens22f1np1: <BROADCAST,MULTICAST> mtu 1500 qdisc noop state DOWN group default qlen 1000
    link/ether a1:b2:c3:d4:e5:c7 brd ff:ff:ff:ff:ff:ff
4: ens33f0np0: <BROADCAST,MULTICAST> mtu 1500 qdisc noop state DOWN group default qlen 1000
    link/ether a1:b2:c3:d4:e5:d6 brd ff:ff:ff:ff:ff:ff
5: ens33f1np1: <BROADCAST,MULTICAST> mtu 1500 qdisc noop state DOWN group default qlen 1000
    link/ether a1:b2:c3:d4:e5:d7 brd ff:ff:ff:ff:ff:ff

Once you have determined the names of your interfaces, you can configure interface bonding in the OS.

 

Static IP configuration

Replace the content of /etc/netplan/50-cloud-init.yaml with the following:

network:
    version: 2
    ethernets:
        ens22f0np0:
            match:
                macaddress: a1:b2:c3:d4:e5:c6
        ens22f1np1:
            match:
                macaddress: a1:b2:c3:d4:e5:c7
        ens33f0np0:
            match:
                macaddress: a1:b2:c3:d4:e5:d6
        ens33f1np1:
            match:
                macaddress: a1:b2:c3:d4:e5:d7
    bonds:
        bond0:
            # MAC address of the server's main public interface
            macaddress: a1:b2:c3:d4:e5:c6
            accept-ra: false
            addresses:
                - 203.0.113.1/32
                - 2001:db8:1:1b00:203:0:112:0/56
            routes:
                - on-link: true
                  to: default
                  via: 100.64.0.1
                - on-link: true
                  to: default
                  via: fe80::1
            nameservers:
                addresses:
                - 213.186.33.99
                - 2001:41d0:3:163::1
            interfaces:
                - ens22f0np0
                - ens22f1np1
            parameters:
                mode: 802.3ad
                lacp-rate: fast
                transmit-hash-policy: layer3+4
        # Optional: private bond configuration
        bond1:
            # MAC address of the first private interface
            macaddress: a1:b2:c3:d4:e5:d6
            accept-ra: false
            interfaces:
                - ens33f0np0
                - ens33f1np1
            parameters:
                mode: 802.3ad
                lacp-rate: fast
                transmit-hash-policy: layer3+4

DHCP configuration

Replace the content of /etc/netplan/50-cloud-init.yaml with the following:

network:
    version: 2
    ethernets:
        ens22f0np0:
            match:
                macaddress: a1:b2:c3:d4:e5:c6
        ens22f1np1:
            match:
                macaddress: a1:b2:c3:d4:e5:c7
        ens33f0np0:
            match:
                macaddress: a1:b2:c3:d4:e5:d6
        ens33f1np1:
            match:
                macaddress: a1:b2:c3:d4:e5:d7
    bonds:
        bond0:
            # MAC address of the server's main public interface
            macaddress: a1:b2:c3:d4:e5:c6
            accept-ra: false
            dhcp4: true
            addresses:
                - 2001:db8:1:1b00:203:0:112:0/56
            routes:
                - on-link: true
                  to: default
                  via: fe80::1
            nameservers:
                addresses:
                - 2001:41d0:3:163::1
            interfaces:
                - ens22f0np0
                - ens22f1np1
            parameters:
                mode: 802.3ad
                lacp-rate: fast
                transmit-hash-policy: layer3+4
        # Optional: private bond configuration
        bond1:
            # MAC address of the first private interface
            macaddress: a1:b2:c3:d4:e5:d6
            accept-ra: false
            interfaces:
                - ens33f0np0
                - ens33f1np1
            parameters:
                mode: 802.3ad
                lacp-rate: fast
                transmit-hash-policy: layer3+4

 

Applying the configuration

The netplan try command can't be used when configuring bonds.

Apply the configuration using the following command:

sudo netplan apply

It may take several seconds for the bond interfaces to come up.

 

Go further

Configuring OVHcloud Link Aggregation in the OVHcloud Control Panel

How to configure OVHcloud Link Aggregation in Windows Server 2019

How to configure OVHcloud Link Aggregation in Debian 9 to 11

How to configure Link Aggregation with LACP in Debian 12 or Ubuntu 24.04

For more information and tutorials, please see our other Dedicated Servers support guides or explore the guides for other OVHcloud products and services.

Related articles